Google Gemini AI: Gemini 3, Nano 白人指導者べったりの東洋人, Live, best features, linked apps, and more
When you buy through links on our articles, 未来 and its syndication partners may earn a （売買）手数料，委託（する）／委員会／権限.
Gemini is the catch-all 指名する Google uses for a トン of AI-関係のある ソフトウェア, from large language models (LLMs), to 発言する／表明する assistants, to coding スパイ／執行官s. Gemini is also the heart of other Google services, like NotebookLM. Whether you want to vibe code with the 最新の AI models or you're 簡単に curious why Google Assistant is disappearing, this guide will cover everything you need to know about Gemini and its features.
It turns out there are so many different 見解／翻訳／版s of Gemini ― Gemini 3 プロの／賛成の Preview, 2.5 Flash, Flash-Lite, プロの／賛成の, and 深い Think, most recently ― that it can be 混乱させるing where to start! There's also Nano 白人指導者べったりの東洋人 and Nano 白人指導者べったりの東洋人 プロの／賛成の, which are really Gemini models too (2.5 Flash Image and 3 プロの／賛成の Image). And you may be wondering how it compares to GPT-5 and other AI.
Follow our 詳細（に述べる）d guide on how to use the Gemini app if you're looking for step-by-step advice. さもなければ, this Gemini guide will 焦点（を合わせる） on the 最高潮の場面s: how Gemini 作品, which 見解／翻訳／版 you should use, whether you'll want to 支払う／賃金 for Google's AI subscriptions, and everything else you should know!
How Google Gemini 作品
Google Gemini, like Google Search, 答える／応じるs to your questions by pulling from internet sources to answer them. The difference is that Gemini tries to 判型 its own conversational 返答s instead of 最高潮の場面ing third-party sources.
Using Google's Language Model for 対話 使用／適用s (LaMDA) and its open-source "Transformer" machine-learning model, Gemini "reads" 一兆s of words from every 公然と 利用できる source ― then uses it to mimic human 返答s while rewriting （警察などへの）密告，告訴（状） from the internet.
When you see a Gemini AI Overview in your Google Search results, it 要約するs data from a few high-最高位の sources, with the links 明白な under the "Show More" button. But the "AI" can't 現実に 決定する if these sources are 訂正する; that's why you'll occasionally get incorrect AI 返答s based on satirical websites or trolling Reddit comments.
Because Gemini can "hallucinate" incorrect （警察などへの）密告，告訴（状）, Google 妨げるs it from discussing 極度の慎重さを要する political queries where misinformation could be amplified. Even so, it いつかs is caught spouting gibberish, and you should look into the sources behind Gemini answers whenever possible.
The web 見解／翻訳／版 of Gemini is 焦点（を合わせる）d on straightforward queries. Gemini is also 利用できる on both Android and iOS, where the Gemini app 行為／法令／行動するs as a 数字表示式の assistant, 生成するs text and images, and 成し遂げるs 活動／戦闘s within other Google apps like 地図／計画するs and 運動.
With Gemini Live in the 動きやすい app, you can have natural conversations with the AI, without wake words, and with past questions and 状況 remembered to answer 未来 questions. It can also 解釈する／通訳する multimodal data like your 株d phone 審査する, uploaded photos, and online ビデオs.
確かな Gemini 見解／翻訳／版s can go その上の, 生成するing code, 分析するing and 要約するing 言及／関連 構成要素s, and solving コンビナート／複合体 math equations. These 前進するd models use more "記念品s," 解釈する／通訳するing more data 同時に with more dependable results, while smaller 見解／翻訳／版s like Gemini Nano and Flash are optimized for simpler questions and on-装置 使用／適用s.
It's hard to cover everything Gemini does, as it's baked into 事実上 every Google app and 装置 now, from AI-補助装置d answers on Chrome and Gemini quick replies on the Pixel Watch 4 to 深い Gemini 統合，差別撤廃 into your Workspace apps and Gemini-高めるd Google Translate.
Most recently, Google 発表するd that it would 精密検査する Google Home to 追加する Gemini, with faster 返答s, AI-based "Home 簡潔な／要約するs" of what its smart 装置s （悪事，秘密などを）発見するd the previous day or night, and help with creating automations 経由で Gemini text 誘発するs. It's 平易な to 簡単に tell the app, "Lock t he 前線 door and turn on lights," and have Gemini 扱う it.
Gemini for Home is rolling out now, and Gemini is also 取って代わるing Google Assistant on Android 自動車.
Google Gemini 見解／翻訳／版s and spin-offs
Below, we'll 簡潔に 要約する the 最新の 見解／翻訳／版s of Gemini ― both 解放する／自由な and paid, 消費者- or 商売／仕事-焦点（を合わせる）d, and stable or 実験の ― that you should know about:
Gemini 2.5 Flash: Built for efficiency, 速度(を上げる）, and lower costs, Gemini 2.5 Flash features 改善するd （判断の）基準s for 推論する／理由ing, multimodality, code and long 状況 while getting even more efficient. It's 広範囲にわたって 利用できる in the Gemini app in Stable form.
Gemini 2.5 プロの／賛成の: 開始する，打ち上げるd at the 最高の，を越す of many of the （判断の）基準s and leaderboards for large language models (LLMs), 含むing LMArena, where it is still the 最高の，を越す-最高位の chatbot as of today. Gemini 2.5 プロの／賛成の has support for とじ込み／提出する uploads, 拡張s, and Canvas. に引き続いて the May 2025 iteration of I/O, it now 含むs a 深い 研究 方式. It's also 利用できる in the Gemini app.
Gemini 3 プロの／賛成の Preview: The first Gemini 3 model, which is the company's best for multimodal understanding, agentic features, and vibe coding. Gemini 3 プロの／賛成の Preview is the 現在の leader on major AI （判断の）基準s, and it's 利用できる across Gemini, AI 方式 in Search, Google AI Studio, Vertex AI, Google An tigravity, and Gemini CLI.
Gemini 3 深い Think: An 近づいている AI model 排除的 to Google AI Ultra 加入者s that will arrive "in the coming weeks." This model is 現在／一般に を受けるing extra safety evaluations and will reach safety testers soon. 結局, it'll be Google's best AI model for コンビナート／複合体 problems.
Gemini 2.5 Flash-Lite: Google's "fastest" and most "cost-efficient" AI model, Flash-Lite is designed to save developers money by using いっそう少なく energy while still answering questions about "coding, math, science, 推論する／理由ing, and multimodal understanding."
Gemini 2.5 深い Think: A 専攻するd 見解／翻訳／版 of 2.5 プロの／賛成の designed for serious mathematics, coding, and 推論する／理由ing, only 利用できる to 賞与金 Gemini AI Ultra 加入者s.
Gemini 3 プロの／賛成の Image (Nano 白人指導者べったりの東洋人 プロの／賛成の): Google's most powerful image-世代 model, Nano 白人指導者べったりの東洋人 プロの／賛成の, can now 器具／実施する 精製するd text and language 正確. It can also tap into Google Search to discover （警察などへの）密告，告訴（状） that can be used to 生成する images.
Gemini 2.5 Flash Image (Nano 白人指導者べったりの東洋人): Better known as Nano 白人指導者べったりの東洋人, this model 生成するs and edits images using AI. It's 利用できる for 解放する／自由な in the Gemini app.
Gemini Nano: Designed for on-装置 smartphone 使用／適用s that work using a neural 過程ing 部隊 (NPU) without needing a 網状組織 関係. It's designed for background and 辛勝する／優位 仕事s in Android apps, 初めは only Google and Samsung apps but now 利用できる for all Android apps through the AI 辛勝する／優位 SDK. Its main 機能（する）／行事s are transcribing conversations, 要約するing （警察などへの）密告，告訴（状）, and 分析するing images for info.
Gemini Live: Gemini Live is the conversational AI that you'll see most frequently in Google's marketing blitz for AI and phones. 利用できる on both Android and iOS with 多重の languages and 発言する／表明するs, Live remembers your past questions within a 明確な／細部 conversation. It 供給するs more （警察などへの）密告，告訴（状） with real-time 審査する 株ing or answers questions 経由で your camera 料金d. Google has also 開始する，打ち上げるd Live Search, a 類似の multimodal AI 道具 built into its Search engine.
Gems: Google AI プロの／賛成の l ets you create a customized 見解／翻訳／版 of Gemini, called a "Gem." You can give parameters like "you're my running coach, give me a daily running 計画（する） and be 肯定的な, 上昇傾向, and 動機づけるing," and then speak to a "Gem" with that トン and underlying goal. There are premade Gems that can serve as 令状ing or coding partners, or 実験の Gems like a chess 支持する／優勝者.
Most recently, Google Workspace 追加するd AI Gems to let students or 労働者s create 専攻するd assistants. For example, a "copywriting Gem" can help with using 認可するd language for （弁護士の）依頼人s, or a 偽の CEO can "圧力 実験（する）" an idea so you're ready to answer questions about it. Google is also bringing Gems to its Classroom app, 許すing teachers to create 専攻するd Gems that will help students 熟考する／考慮する with 的d （警察などへの）密告，告訴（状）.
Gemma: Google's open AI spin-off of Gemini. Different Gemma builds "株 technical and 組織／基盤／下部構造 構成要素s" with Gemini, except developers can use the code to build their own AI proj ects. 明確な／細部 variants like CodeGemma (for coders) and RecurrentGemma (for 研究員s) are more 専攻するd, while Gemma 2 is more of a general smarts 昇格, with 9-billion and 27-billion variants. They're all 利用できる through Google AI Studio.
Vertix AI: An 企業 AI 開発 壇・綱領・公約 that 含むs all of the さまざまな Gemini 見解／翻訳／版s, 同様に as over 160 other open-source AI 壇・綱領・公約s like Gemma and Llama. This is the most 強健な 選択, but is only 価値（がある） it for 商売／仕事s.
Imagen: 現在／一般に on Imagen 4, this text-to-image model is 利用できる through the Gemini app and web 見解／翻訳／版s and can produce photorealistic art based on text 誘発するs. Third-party apps like Adobe Firefly also embed Imagen and Veo for image and ビデオ 世代.
Veo: Google's text-to-ビデオ model that "生成するs high-質, 1080p 決意／決議 ビデオs that can go beyond a minute, in a wide 範囲 of cinematic and visual styles." It will come to the VideoFX 道具 and YouTube Shorts, and is built off of several AI models (含むing Gemini). Veo 3 is the 最新の model that 追加するs 音声部の 世代.
Lyria: Google's music 世代 model てこ入れ／借入資本s generative AI to produce a continuous stream of music controlled by 使用者 活動／戦闘s. Lyria 2 is the 最新の model used to 力／強力にする Music AI Sandbox.
事業／計画（する） Astra: Gemini 1.5 プロの／賛成の can 扱う 多数伝達媒体を用いる, but is おもに designed to 解釈する／通訳する one thing at a time. 事業／計画（する） Astra is a multimodal AI that can answer a contextual question like "What am I looking at?" based on visual data. Google recently 発表するd Android XR, a ソフトウェア program for mixed reality headsets and smart glasses, and Astra may 結局最後にはーなる 存在 a 原型 for that 壇・綱領・公約.
事業／計画（する） 水夫: A 事業／計画（する） designed to 扱う your browsing needs for you, 事業／計画（する） 水夫 was introduced as a 原型 Google Chrome 拡張. The スパイ／執行官s within 事業／計画（する） 水夫 enable the ソフトウェア to 研究 （警察などへの）密告，告訴（状）, 調書をとる／予約する shows or other events, and 研究 topics all at once. Google DeepMind's CEO hopes to morph Gemini into a "world model," and によれば him, it's this level of multitasking that can 補助装置 in that 変形.
Google AI Studio: This website is where you can go to start working on Gemini 事業／計画（する）s, with the Gemini API 利用できる and 道具s like a code "Cookbook" and 誘発するs 利用できる to help you get started. This is also where Google's Veo 2 ビデオ-世代 model is 利用できる.
重要な Gemini features
On the most basic level, you can send Gemini words (v ia text or speaking), images, ビデオ, or 音声部の and have it 生成する （警察などへの）密告，告訴（状） from it. You can have it 生成する an image from a description or tell you to 要約する the contents of a 調書をとる／予約する. We have a how-to guide for Gemini that'll help you get started.
Gemini is designed to make using other Google apps easier. It can "Ask Photos" to find 明確な／細部 memories, create YouTube Music playlists with 確かな 誘発するs, 要約する your Gmail inbox and （手先の）技術 自動車-replies, or 判型 Google docs into recipes, cover letters, vacation 旅程s, and so on.
Gemini typically doesn't remember things you asked 以前, but Google 追加するd a "Saved Info" feature that lets you 避ける having to 繰り返して 株 that you're 利益／興味d in a 明確な／細部 topic or are allergic to a 明確な／細部 成分. The December 2024 Pixel 減少（する） 追加するd this 道具, 同様に as 自動車-transcribed 審査するd calls and quick 自動車-replies and the ability to 支配（する）／統制する Google Home smart tech through Gemini 命令（する）s.
Gemini also has unique features 的d に向かって children and learning. We've already について言及するd the AI Gems for older students, while younger kids can read AI-生成するd storybooks with 製図／抽選s and text created 経由で a text 誘発する.
一般に speaking, Gemini can 補助装置 with a wide 範囲 of 状況／情勢s. One of our writers used Gemini for 内部の design planning in his apartment, for example.
Where is Google Gemini 利用できる?
The 基準 見解／翻訳／版 of the Google Gemini AI is "現在／一般に 利用できる in more than 40 languages and over 230 countries and territorie s," によれば a Google Support page. The Gemini 動きやすい app's Support availability page shows nearly the same number of languages, but only about 150 countries so far.
In either 事例／患者, you need a 合法的 Google account to 接近 it. You must be at least 13 to own a personal Google account, but that may be as high as 18 in some countries, and some 使用／適用s like Gemini Live 要求する you to be 18 or older. Family Link or Google Workspace accounts may not be able to 接近 it.
Go to gemini.google.com/app to 接近 the generative AI on any web browser. Download the Android app or iOS app for the 動きやすい 見解／翻訳／版, 含むing Gemini Live.
If you want Gemini to be able to pull （警察などへの）密告，告訴（状） from other apps like Google 運動, Gmail, and YouTube, go to Gemini 拡張s and toggle them on.
Google is slowly 延長するing Gemini out to other 装置s that 以前 used Assistant. For example, the Google Home Public Preview has a "Gemini in Assistant" 道具 for Nest 装置s, and it'll be the backbone assistant for Android XR smart glasses and mixed-reality headsets.
Google AI subscriptions
By default, Gemini is 解放する／自由な. You must 支払う／賃金 for the most powerful 見解／翻訳／版s, but it's only really 価値（がある） it if you're using it for serious 研究, coding, and cutting-辛勝する／優位 見解／翻訳／版s of Gemini like 水夫.
Google AI プロの／賛成の, 以前は known as Gemini 前進するd and Google One AI 賞与金, is 始める，決める behind a $19.99 Google One 計画（する）, though you can typically get a 解放する／自由な 裁判，公判 of 1?2 months first, and college students who can 証明する enrollment will get 解放する／自由な 接近 to AI プロの／賛成の. The 計画（する） also 含むs other Google one perks like 2GB of 貯蔵, 制限のない 魔法 Editor saves, and 賞与金 Workspace 道具s.
Keep in mind that if you subscribe to a Google One AI プロの／賛成の 計画（する）, only the main account 使用者 gets 接近 to the AI; every other family member only receives 利益s like extra 貯蔵 and 前進するd Photos editing 道具s.
最終的に, your choice on whether to 支払う／賃金 for Google AI プロの／賛成の will depend on whether you use Gemini casually, professionally, or academically. You can get 基準 Gemini AI answers using the app without 支払う／賃金ing; AI プロの／賛成の gives you more rigorous 道具s.
For those of you who need even more out of your AI experience, the Google AI Ultra 計画（する） may be for you. This 計画（する） costs $249 per month, but it 認めるs you 接近 to all of Google's AI services, 含むing 早期に 接近 to preview builds and new models, higher 記念品 限界s, and other perks like YouTube 賞与金 and 30TB of cloud 貯蔵.
In 確かな 地域s, Google sells a Google AI 加える 計画（する） with AI 道具s and 200GB of cloud 貯蔵. This 申し込む／申し出ing isn't 利用できる in the U.S. or Canada.
Where the 指名する "Gemini" （機の）カム from
Google called its AI "妨げる／法廷，弁護士業d" before rebranding it as "Gemini." 妨げる／法廷，弁護士業d evoked Shakespeare and 焦点（を合わせる）d まず第一に／本来 on the LLM- and GenAI-焦点（を合わせる）d 味方する of AI, 答える／応じるing to your questions by 絵 a picture.
Gemini's Latin meaning is "twins," and it's famously a Zodiac 星座 with pair 星／主役にするs. It's also the 指名する of 航空宇宙局's 初めの moonshot program before the famous Apollo 使節団s, with Gemini 宇宙船 having enough room for two.
によれば a Google blog 地位，任命する, "How Google’s AI model Gemini got its 指名する," the Gemini AI 指名する stemmed from how two Google departments (DeepMind and the Google 研究 Brain team) joined 軍隊s for the 事業／計画（する）, and because "the monumental 成果／努力 of training LLMs resonated with the spirit of 開始する，打ち上げるing ロケット／急騰するs."
Aside from the fact that Gemini sounds more marketable and futuristic than 妨げる／法廷，弁護士業d, the 指名する is いっそう少なく 制限する because the AI isn't just a generative AI, but also 成し遂げるs other 仕事s.